Rice - ECE Affiliates Posters 2003

Department of Electrical and Computer Engineering
Rice University
Affiliates Meeting - Poster Titles
Wednesday, October 8, 2003

The Connexions Project - cnx.rice.edu
ELEC424/427: High Speed & Embedded Systems Design I/II: A Systems Design & Lab Course Based on the Texas Instruments TMS320F2812 DSP - Frantz* and Fernandez
New Multivariate Dependence Measures and Applications to Neural Ensembles - Goodman and Johnson*
An Information Processing Approach to Distributed Detection - Lexa and Johnson*
Multiscale connection-level analysis of network traffic - Sarvotham, Riedi*, and Baraniuk*
Multiscale Geometry Model for Image Processing - Wakin, Romberg, Choi*, and Baraniuk*
Digital Signal Processing for Communication Systems - Ahmed and Baraniuk*
A New Interpretation of Translation Invariant Image Denoising - Hua and Orchard*
Complex Image Representations Based on Single Quadrant Spectrum of Wavelet Subbands - Ates and Orchard
pathChirp: Efficient Bandwidth Estimation for Network Paths - Ribeiro, Riedi*, and Baraniuk*
Safari: A Scalable Architecture for Ad Hoc Networking and Services - Post, Chaudhuri, Saha, Shu, Ahamed, Riedi*, Druschel*, Johnson*, Baraniuk*, and Hu*
Spinach: A Simulator for Programmable Network Interface Architectures - Willmann, Brogioli, and Pai*
The Embedded Triangles Algorithm for Distributed Estimation in Sensor Networks - Delouille, Neelamani, Chandrasekaran, and Baraniuk*
GNOMES: A Testbed for Low Power Heterogeneous Wireless Sensor Networks - Welsh, Ganier, Cheng, Kriengchaipruck, Yin, Valenzuela, Fish, and Frantz*
Wireless Camera Network Calibration - Mantzel, Baraniuk* and Choi*
Little Lasers Weather Sensory Network - Davenport, Printy, Dhara, Harnoy, Lai, Young*, Baraniuk*, and Frantz*
High Performance MPI Libraries with TCP/IP - Majumder, Rixner*, and Pai*
Design, Analysis, and Implementation of DVSR: A Fair, High Performance Protocol for Packet Rings - Gambiroza, Yuan, Balzano, Liu, Sheafor, and Knightly*
Challenges in enabling large-scale wireless broadband - Chawathe, Robinson, Karrer, and Knightly*
Implementation Challenges of a Wide Area TAP Network - Aazhang*, Sabharwal*, Frantz*, Murphy, and Lou
Multilinear Algebra for Signal Processing - de Baynast and Aazhang*
Implementation of MIMO Equalizers for Mobile Handsets - Radosavljevic, de Baynast, and Cavallaro*
Implementation of LDPC Encoder/Decoder - Karkooti and Cavallaro*
Reconfigurable Stream Processors - Rajagopal, Rixner*, and Cavallaro*
A Reconfigurable Gigabit Ethernet NIC - Mohsenin, Swanson, Frantz*, and Rixner*
The Propagation of Terahertz Pulses Through Inhomogeneous Media - Pearce, Jian, and Mittleman*
Spatial correlations in time-resolved terahertz speckle patterns - Jian, Pearce, and Mittleman*
Sub-wavelength Resolved Terahertz Imaging by using Near-field Optical Antennas - Wang, Barkan, and Mittleman*
Sensitive Measurement of Carbonyl Sulfide with a Thermoelectrically Cooled Quantum Cascade Laser: Application in Medical Diagnostics - So, Wysocki, Roller, Kosterev, Tittel*, Curl, Bag, Paraguya, Gmachl, and Sivco
Breath Diagnostics: Off-Axis Integrated Cavity Output Spectroscopy for Nitric Oxide Detection in Human Breath using a Quantum Cascade Laser - McCurdy, Bakhirkin, Kosterev, Curl, Allen and Tittel*
Ultrafast All-Optical Switching using the Dynamic Franz-Keldysh Effect - Srivastava, Srivastava, Wang, and Kono*
Ultrafast Photoinduced Softening in a Magnetic Semiconductor - Wang, Khodaparast, Kono*, Slupinski, Oiwa, and Munekata
Investigating Carbon-based Nanoelectronics - Chu and Kelly*
Single-molecule Imaging and Manipulation - Osgood and Kelly*
Atomic-level Investigation of Fluorinated Carbon Nanotubes - Takhar and Kelly*
Finite difference time domain studies of optical properties of nanoshell structures - Nordlander* and Oubre
Surface Enhance Raman Scattering (SERS) and Surface Plasmon Resonance (SPR) Sensing with Metal Nanoshells - Jackson, Tam, Mirin, and Halas*
Plasmonic Properties of Metallodielectric Periodic Structures - Steele, Moran, and Halas*
Rice ECE Activities in the Medical Center: Multiphoton Microscopy and Spectroscopy - Iyer, Rossow, Losavio, Waxham, Saggau*
A Model of the Respiratory Central Pattern Generator in Rat - Amini, Zwischenberger, Bidani, Byrne, and Clark*
Vascular Smooth Muscle Relaxation: Modeling the Nitric Oxide/cGMP Pathway - Yang and Clark*

The Connexions Project
Connexions is a collaborative, community-driven approach to authoring, teaching, and learning that aims to capture and convey the dynamic, continuum of knowledge by fully exploiting information technology. Available free of charge to anyone under open-content and open-source licenses, Connexions offers high-quality, custom-tailored electronic course material, is adaptable to a wide range of learning styles, and provides tools that encourage students to explore the links among concepts, courses, and disciplines. In contrast to the traditional process of textbook writing and publishing, Connexions fosters world-wide, cross-institution communities of authors, instructors, and students, who collaboratively and dynamically fashion "modules" from which courses are constructed.
Launched in 1999 at Rice University and supported so far by $2m in start-up funding from Rice University and the Hewlett Foundation, Connexions has produced a free, open-source toolset for authoring modules, building courses, and navigating and exploring modules. Connexions is also fostering worldwide, cross-institutional communities of authors, instructors, and students. In particular, we have built a strong and growing community-of-use in the electrical and computer engineering area of digital signal processing (DSP) -- participating institutions currently include Rice University, University of Illinois, Georgia Tech, the Ohio State University, Polytechnic University, Cambridge University, and Technical University of Norway. A project with UNESCO, MIT, and Carnegie Mellon University will see Connexions support content development and use in a number of developing countries.
ELEC424/427 - High Speed & Embedded Systems Design I/II: A Systems Design & Lab Course Based on the Texas Instruments TMS320F2812 DSP
In today's global engineering market, it is critical for graduates to receive a balanced education in both the theoretical and practical aspects of DSP and design. Rice University's ECE Department is redeveloping a fall/spring course sequence that will serve to educate students in the implementation of DSP and embedded applications. The course is aimed at upper level students with previous experience in an embedded environment and basic knowledge of DSP theory. In the fall semester, students will be given and learn to program custom-designed F2812-based hardware. This platform is meant to give students wide exposure to sensing, signal processing, wireless communications, and networking applications. In the spring semester, students will design and fabricate their own add-on PCB to be tested on the F2812 hardware. At the end of each semester, students will write an application note that will be submitted to TI. All course content will be hosted by the Connexions project, which facilitates future use of this course material by others.
New Multivariate Dependence Mesures and Applications to Neural Ensembles
We develop two new multivariate statistical dependence measures. The first, based on the Kullback-Leibler distance, results in a single value that indicates the general level of dependence among the random variables. We show how this measure can be decomposed into components representing interactions of different orders. The second, based on an orthonormal series expansion of joint probability density functions, provides more detail about the nature of the dependence. We show how this measure can be used to decompose Pearsonâs phi-squared measure of association. We apply these dependence measures to the analysis of simultaneous recordings made from multiple neurons, in which dependencies are time-varying and potentially information bearing.
An Information Processing Approach to Distributed Detection
We apply the recent theory of information processing to a hybrid distributed detection architecture that combines the traditional parallel and tandem architectures. Central to this theory is the Kullback-Leibler discrimination distance and a quantity known as the information transfer ratio, defined as the ratio of the KL distances between the distributions characterizing the input and output of a system. We characterize the asymptotic performance of a proposed hybrid system and compare it with the performance of the parallel, tandem and centralized architectures.
Multiscale connection-level analysis of network traffic
Network traffic exhibits drastically different statistics, ranging from nearly Gaussian marginals and long range dependence at very large time scales to highly non-Gaussian marginals and multifractal scaling on small scales. This behavior can be explained by decomposing traffic into two components according to the connection bandwidth: the small bandwidth component absorbs most traffic and is Gaussian, while large bandwidth component constitutes virtually all of the small scale bursts. Based on this understanding, we propose a novel connection-level traffic model that parsimoniously accounts for user behavior, network topology, and the heterogeneous distribution of network bandwidths.
Multiscale Geometry Model for Image Processing
Natural images can be viewed as combinations of smooth regions, textures, and geometry. The wavelet transform, popular in many standard image processing algorithms, provides reasonably efficient representations for smooth and texture features, but not for geometric features such as edges. As a result, most wavelet-domain algorithms fail to accurately model the coefficients in geometric regions. In this poster, we present a novel Muliscale Geometry Model (MGM) for geometric features that can be applied to a variety of transforms and applications. With a real-valued wavelet transform, the MGM allows us to develop an image coder with efficient descriptions of geometric wavelet coefficients. With a complex-valued (redundant) wavelet transform, the MGM allows us to explicitly capture the multiscale behavior of coefficient magnitudes and phases; we demonstrate the effectiveness of this model by applying it to feature extraction.
Digital Signal Processing for Communication Systems
We discuss some of the advances made by the DSP group in the ECE department at rice in the area of communication systems.
Broadband access via digital subsriber lines (DSL) has seen explosive growth in the past few years. Crosstalk is the dominant source of interference in such systems and severely affects achievable data rates for such systems. We present both crosstalk avoidance and crosstalk cancellation techniques that significantly improve the performance of DSL systems.
Delay-constrained communications systems over fading channels is of particular interest to engineers as they accurately model real-world wireless systems. Often in such systems the hostile nature of the channel makes the effective data rate, or throughput, much lower than the attempted transmission rate. Using knowledge of the channel statistics, we present signal processing techniques that use optimal rate and power control to maximize the throughput. The gains are very significant and lead to some rather striking conclusions that are counter to conventional wisdom in the area.
A New Interpretation of Translation Invariant Image Denoising
Translation Invariant (TI) image denoising is a frame denoising method. It outperforms orthogonal wavelet thresholding by averaging a collection of image estimates in different orthogonal bases. In this poster, we propose local models for characterizing the statistics (bias and variance) of this collection of estimates. Using these models, TI's gain can be analyzed due to the convexity of the error metric (MSE). Motivated by the edge geometry and analysis of smooth regions, we design a special way to choose an estimate (accordingly the basis generating it) from the collection at each pixel. In this way, the visual quality improvement of TI can also be explained. Insights drawn from this perspective include: a) the mechanisms by which TI achieves gain are different in smooth and edge regions, and b) most gain comes from edge regions. We also point to an improved way of exploiting the statistics mentioned above, if the position information of edges is available.
Complex Image Representations Based on Single Quadrant Spectrum of Wavelet Subbands
High frequency structures, such as edges, texture, contain significant portion of the information available in images. For coding purposes, the image representation has to provide direct access to this information content without introducing any redundancy. We introduce here such a compact representation based on complex signals in wavelet domain, whose spectrum has only one nonzero quadrant. The corresponding magnitude and phase responses are linked with the most significant edge properties, namely energy, location and orientation. Simulations show that phase responses provide unbiased estimates of the edge location and orientation. We discuss potential applications of this representation in various fields including image coding.
pathChirp: Efficient Available Bandwidth Estimation for Network Paths
pathChirp is a new active probing tool for estimating the available bandwidth on a communication network path. Based on the concept of "self-induced congestion," pathChirp features an exponential flight pattern of probes we call a chirp. Packet chirps offer several significant advantages over current probing schemes based on packet pairs or packet trains. By rapidly increasing the probing rate within each chirp, pathChirp obtains a rich set of information from which to dynamically estimate the available bandwidth.
Safari: A Scalable Architecture for Ad Hoc Networking and Services
The wide spread usage of powerful mobile computing devices predict sizes for future Mobile Ad hoc Networks (MANETs), several magnitudes larger than current protocols can handle. Our research is targeted towards providing scalable ad hoc routing and enable conventional internet services like the DNS, DHCP and SMTP for such MANETs and eventually seamless integration of wired infrastructure, whenever available. To this end we have proposed an architecture called Safari which employs topology aware, hierarchical addressing for the mobile hosts through a proactive, self-organizing, hierarchical address assignment protocol. Topology awareness is implemented in a distributed fashion using a Distributed Hash Table (DHT). Novel to our approach are beacon broadcasts which are instrumental in creating an adaptive hierarchy and an overlay structure, and at the same time disseminate valuable routing information, thus drastically reducing overhead.
Spinach: A Simulator for Programmable Network Interface Architectures
Spinach is a modular architectural simulator designed to efficiently explore the design space of programmable network interface architectures. Spinach models generic system components (e.g., ALUs, control paths, instruction processing), as well as those specific to embedded systems and network interfaces (e.g., software-controlled scratchpad memory, hardware assists for DMA and medium access control.) Spinach's modularity allows wide design space exploration with little or no code variation.
The Embedded Triangles Algorithm for Distributed Estimation in Sensor Networks
We propose a new iterative distributed estimation algorithm for Gaussian hidden Markov graphical models with loops. We decompose a loopy graph into a number of linked {\em embedded triangles} and then apply a parallel block-Jacobi iteration comprising local linear minimum mean-square-error estimation on each triangle (involving a simple $3\times 3$ matrix inverse computation) followed by an information exchange between neighboring nodes and triangles. A simulation study demonstrates that the algorithm converges extremely rapidly, outperforming a number of existing algorithms. Embedded triangles are simple, local, scalable, fault-tolerant, and energy-efficient, and thus ideally suited for wireless sensor networks.
GNOMES: A Testbed for Low Power Heterogeneous Wireless Sensor Networks
The design of low-power, small form-factor remote and mobile sensing systems has become a more feasible task in the past few years due to several continuing trends. The cost of solid-state sensors for a wide variety of applications keeps decreasing. Robust low-power and short-range radio hardware has emerged which can handle moderate to high data rates (approx. 1Mbit/s). Embedded microprocessors consume much less power than their predecessors while achieving much better levels of performance. All of these trends make feasible very dense networks of fixed and mobile wireless devices for use in many different sensing and decision-making systems. In this poster we present a low-cost hardware and software testbed, named GNOMES, that has been developed at Rice University to explore the properties of heterogeneous wireless sensor networks, In particular, we present various methods to extending the lifetime of individual nodes in the network, the design tradeoffs that this presents, and the impact that this will have on the performance of the sensor network.
Wireless Camera Network Calibration
Sensor networks show potential to estimate complex spatial systems efficiently through distributed computing power and wireless communications. As the microprocessors on these nodes advance, it will become feasible to operate with multi-dimensional data such as video streams to learn even more about the environment around the network. The non-mobile camera nodes of this network must first be calibrated and their positions and orientations estimated. Then, an observed point or event can be mapped to a specific direction, or ray of light from the point to the camera. This opens up many applications in image based rendering techniques and shape recovery.
Little Lasers Weather Sensory Network
Using mere laser pointers, we are building a low-cost, low power weather sensory network to achieve high resolution weather data. We can calculate the path-averaged rain rate over a distance by measuring the variation in the intensity of the received laser beam. Further areas of research include measuring wind speeds and circular winds.
High Performance MPI Libraries with TCP/IP
Cluster computing relies heavily on message passing and libraries developed for this purpose. However one major drawback of the current crop of MPI libraries is that they do not utilize the full potential offered to them by gigabit ethernet. The bandwidth achieved by these libraries tends to saturate at a much lower value than the theoretical line rate of around 940 Mbps. Our research is focussed on improving this performance of MPI libraries. We use TCP/IP as the communication protocol underneath the MPI communication primitives. Our modifications to the Los Alamos MPI (LAMPI) library have already shown significant improvement in the bandwidth performance of the library. We have been able to increase the acheivable bandwidth of the LAMPI library from 350 Mbps to 650 Mbps. However there still is room for improvement as the bandwidth is still shy of the line rate and this forms the crux of our current endeavors.
Design, Analysis, and Implementation of DVSR: A Fair, High Performance Protocol for Packet Rings
The Resilient Packet Ring (RPR) IEEE 802.17 standard is a new technology for high-speed backbone metropolitan area networks. A key performance objective of RPR is to simultaneously achieve high utilization, spatial reuse, and fairness, an objective not achieved by current technologies such as SONET and Gigabit Ethernet nor by legacy ring technologies such as FDDI. The core technical challenge for RPR is the design of a bandwidth allocation algorithm that dynamically achieves these three properties. The difficulty is in the distributed nature of the problem, that upstream ring nodes must inject traffic at a rate according to congestion and fairness criteria downstream. Unfortunately, we show that under unbalanced and constant-rate traffic inputs, the RPR fairness algorithm suffers from severe and permanent oscillations spanning nearly the entire range of the link capacity. Such oscillations hinder spatial reuse, decrease throughput, and increase delay jitter. In this project, we introduce a new dynamic bandwidth allocation algorithm called Distributed Virtualtime Scheduling in Rings (DVSR). The key idea is for nodes to compute a simple lower bound of temporally and spatially aggregated virtual time using per-ingress counters of packet (byte) arrivals. We show that with this information propagated along the ring, each node can remotely approximate the ideal fair rate for its own traffic at each downstream link. Hence, DVSR flows rapidly converge to their ring-wide fair rates while maximizing spatial reuse. To evaluate DVSR, we develop an idealized fairness reference model and bound the deviation in service between DVSR and the reference model, thereby bounding the unfairness. With simulations, we find that compared to current techniques, DVSR's convergence times are an order of magnitude faster (e.g., 2 vs. 50 msec), oscillations are mitigated (e.g., ranges of 0.1% vs. up to 100%), and nearly complete spatial reuse is achieved (e.g., 0.1% throughput loss vs. 33%). Finally, we provide a proof-of-concept implementation of DVSR on a 1 Gb/sec network processor testbed and report the results of testbed measurements.
Challenges in enabling large-scale wireless broadband
In this poster we describe the key challenges of resource control in a TAP network. A TAP network is a first-of-its kind broadband wireless network that provides high-performance access to the wired Internet to residences and public places in a large area. On top of directional antennas that provide high transmission rates and spatial reuse, we formulate the challenges of controlling resources in a distributed, wireless environment at the MAC layer. We propose distributed, opportunistic and coordinated resource management to achieve system-wide high performance. An opportunistic use of resources is required due to the fast timescale of channel variations, allowing the exploitation of high-quality channels. Distributed resource management allows an efficient use of the resources by forwarding packets along high-quality routes. Finally, resource coordination is needed to ensure a fair resource allotting to multiple users and to exploit spatial reuse.
Implementation of a Wide Area TAP Network
This poster describes the hardware and algorithmic implementation challenges in building a network of wireless Transit Access Points (TAPs). A TAP is a wireless network node equipped with multiple air interfaces, capable of providing high-speed data links to both mobile users and other TAPs. In particular, these TAP-to-TAP links allow an access point to be connected to a larger network without a wired link. The elimination of the need for a wired connection at every access point will significantly reduce the cost and ease the installation of additional access points. We describe here the current efforts to develop custom TAP hardware and to implement the multiple-antenna physical layer algorithms required to provide the high speed services described above. The hardware challenges include designing a system with multiple radio interfaces and providing sufficient resources to implement advanced multiple-antenna algorithms. The algorithmic challenges include designing high throughput, error resistant multiple antenna schemes and implementing them in real hardware.
Multilinear Algebra for Signal Processing
The aim of this work is to generalize concepts from vector and matrix algebra to tensor algebra, and to use them as tools in signal processing. The next generation of wireless systems (3G) will use the code division multiple access (CDMA) technique: all users are simultaneously transmitted in the same frequency bandwidth. In order to separate them at the receiver, an orthogonal spreading sequence is assigned at each user before the transmission. In the case of frequency-selective slow fading channels, most of the receivers requiring statistics estimation of the data, are very sensitive to the channel variations and provide poor performance. In that case, we show that we are able to blindly recover the transmitted data streams, the antenna gains and the channel taps by performing the canonical decomposition of the 3-D data tensor. The uniqueness of the canonical decomposition is guaranteed under mild assumptions: short frame length (5-10 symbol period is enough in most cases), orthogonality between transmitted antennas is not required, overload system is supported.
Implementation of MIMO Equalizers for Mobile Handsets
Use of multiple transmit and receive antennas (MIMO system) is an emerging strategy for increasing the data rate and spectral efficiency in wireless cellular communication systems. The focus of this project is the design and hardware implementation of channel equalizers as the crucial part of the MIMO mobile handset. In a Wideband CDMA downlink, multipath propagation channels change the users' spreading waveforms causing multiple access interference (MAI) between the active users. In order to reduce the MAI in the downlink, linear channel equalization on the mobile side is proposed. A channel equalizer can be viewed as the filter inverse to the channel that is able to restore the orthogonality of the users' spreading codes in the system. The aim of the project is also to design a scalable and programmable architecture for the channel equalizers. This architecture should be able to handle different environments (slow and fast fading channels) and different antenna configurations on the base station and mobile side.
Implementation of LDPC Encoder/Decoder
This research proposes VLSI Architectures for decoding Low Density Parity Check (LDPC) codes. The inherent parallelism and good error correction capability of LDPC codes leads to their very high throughput and near Shannon limit decoding performance. It has been shown that these codes can achieve higher bit error rates than the best-known Turbo codes in low signal to noise ratio conditions. Low power architectures that meet real time data requirements for handset devices are very important for next generation of wireless communication systems.
A (3,6) LDPC decoder with a block length of 1500 bits and rate _ has been implemented using programmable hardware. The decoder is prototyped in both VHDL (hardware description language) code and in LabVIEW FPGA code for the National Instruments reconfigurable hardware. This architecture uses the "Modified Min-Sum" algorithm, an approximation to the sum-product algorithm, which has very good performance with low complexity. A semi-parallel approach is used to balance the tradeoff between performance, area and power consumption.
Reconfigurable Stream Processors
This poster presents the design and use of reconfigurable stream processors for the physical layer of wireless systems. Stream processors, traditionally used for high performance media processing, use clusters of functional units to provide support for hundreds of functional units in a programmable architecture. We provide support for reconfiguration in stream processors, enabling them to be power-efficient by adapting to the compute requirements of the application. We demonstrate the real-time implementation of a 32-user base-station, employing multiuser channel estimation, multiuser detection and Viterbi decoding physical layer algorithms and supporting 128 Kbps/user at an estimated power consumption of 8 Watts at 935 MHz. The reconfigurable stream processor adapts the number of clusters, functional units and frequency dynamically with the workload. When the application workload changes to say 4 users, the reconfiguration support allows the power consumption to reduce by a factor of 10X, from 8 Watts to 800 mW.
A Reconfigurable Gigabit Ethernet NIC
Modern network interfaces provide fixed functionality and are optimized for sending and receiving large packets. Our past research has shown that both increased functionality in the network interface and increased bandwidth on small packets can significantly improve the performance of today's network servers. To address these issues, we are developing a prototype network interface card using the Avnet Virtex-II Pro development board. The Avnet board contains multiple FPGAs, providing flexibility, and multiple memories, including a 128MB SO-DIMM. Preliminary results show that we are able to receive data at line rate and store it in SDRAM for all packet sizes. This is quite promising, since no existing network interface can send or receive small packets at line rate. Ultimately, we intend to use the flexibility of the FPGAs and the memory capacity of the DRAMs to implement specialized services on the network interface to improve overall server performance.
The Propagation of Terahertz Pulses Through Inhomogeneous Media
The propagation of waves through inhomogeneous media gives rise to many interesting but complex phenomena. We describe measurements involving the transmission of single-cycle terahertz pulses through a scattering medium. Using terahertz time-domain spectroscopy, we measure the electric field of a multiply scattered wave with a time resolution shorter than one optical cycle. This time-domain measurement provides information on the statistics of both the amplitude and phase distributions of the diffusive wave. We are able to measure both the unscattered (ballistic) and the multiply scattered (diffuse) portions of the electric field. Additionally, we are able to extract the propagation constants of the medium and compare them to predictions of the quasi-crystalline approximation. We demonstrate the usefulness of the terahertz time-domain spectrometer to investigate scattering.
Sub-wavelength Terahertz Imaging by using Near-field Optical Antennas
One of the most active areas of the current research in terahertz imaging involves improving the resolution below the diffraction limit. At the same time, optical antennas have drawn increasing attention due to their ability to concentrate electromagnetic waves in a sub-wavelength region. By combining apertureless near-field scanning optical microscopy (ANSOM) with terahertz time-domain spectroscopy (THz-TDS), we realize the sub-wavelength resolved THz imaging and demonstrate a spatial resolution of λ/500. We also study the near-field response of the optical antenna and the wave propagation effects on the antenna.
Sensitive Measurement of Carbonyl Sulfide with a Thermoelectrically Cooled Quantum Cascade Laser: Application in Medical Diagnostics
Analysis of expired breath to determine disease states, exposure, and susceptibility of the human body remains an area of research still in its infancy. A lack of available technologies capable of quantifying volatile molecular species in expired human breath with concentrations in pico-molar or low parts-per-billion (ppb) range is the main limiting factor in clinical breath analysis. A relatively new approach to the quantitative analysis of molecular species in expired breath is mid-infrared laser absorption spectroscopy (IRLAS), which is capable of robust and real-time measurements without complex sample preparations. An IRLAS system using thermoelectrically cooled quantum cascade (QC) lasers was developed for the purpose of quantifying carbonyl sulfide (COS). Recent medical literature has shown exhaled COS to be an important biological marker in diagnosing and monitoring liver diseases as well as acute lung transplant rejection. A minimum detection limit of 30 ppb was achieved using a 36-meter optical pathlength. Selectivity of ¹²C¹⁶O³²S and ¹²C¹⁶O³⁴S stable isotopes was also demonstrated and could be useful in identifying the pathologic origin of the sulfur component in the above-mentioned disorders. The results from pilot clinical studies measuring the exhaled breath of lung transplant recipients suggest that the development of medical instrumentation based on QC lasers is feasible.
Breath Diagnostics: Off-Axis Integrated Cavity Output Spectroscopy for Nitric Oxide Detection in Human Breath using a Quantum Cascade Laser
Asthma is a prevalent disease diagnosed by pulmonary function testing. Asthmatic patients have higher concentrations of exhaled nitric oxide (NO) than healthy counterparts [1,2], and detecting exhaled NO may provide a test with higher sensitivity and specificity than current diagnostic methods. Laser absorption spectroscopy in mid-IR region combined with long optical path cell provides the required sensitivity for breath diagnostics such as asthma. A new generation of semiconductor lasers, quantum cascade lasers (QCL) opens the way to realize a compact gas sensor that is compatible with a clinical setting. An effective optical pathlength exceeding several hundred meters can be achieved in a gas cell of just a few cm in physical length by using ultra low-loss dielectric mirrors. The mirrors form an optical cavity, and the intracavity absorption can be measured either through a change of the cavity ringdown time (CRDS) or using an integrated cavity output spectroscopy (ICOS) technique. In this work we investigated the feasibility of an off-axis ICOS approach with a short cavity. The cavity was formed by two concave ultra-loss mirrors (2 inch in diameter, 1 m radius of curvature) separated by 5 cm. The optimum NO detection sensitivity at 1920.7 cm^-1 realized to date is 15 ppb.
Investigating Carbon-based Nanoelectronics
In order to maintain the current trends in miniaturization in the microelectronics industry, new materials and techniques to create and to control the electronic properties of nanometer-scale devices are needed. Two possible alternatives to overcome the limitations in further shrinking of current silicon-based structures include carbon nanotubes and single molecule-based devices. We have begun investigating the electronic properties of both these structures utilizing scanning tunneling microscopy. The information gained from these experiments will be critical in the design and construction of future electronic devices.
Ultrafast All-Optical Switching using the Dynamic Franz-Keldysh Effect
We report the experimental observation of the dynamic Franz-Keldysh effect (DFKE) in GaAs bulk sample at room temperature. The salient features of this electro-optical effect include ultrafast changes in the band structure of the solid leading to absorption below and photoinduced transparency above the band edge. We used ultrashort pulses of long-wavelength light to bring about the above-mentioned changes. In addition, we propose schemes making use of this effect to realize all-optical wavelength conversion required in high-speed optical networks based on wavelength division multiplexing (WDM).
Ultrafast Photoinduced Softening in a Magnetic Semiconductor
We have used two-color time-resolved magneto-optical Kerr Effect (MOKE) spectroscopy to manipulate and detect dynamic processes of spin/magnetic order in a ferromagnetic semiconductor InMnAs. We observed ultrafast photo-induced softening (i.e., transient decrease of coercivity) due to spin-polarized transient carriers. This transient softening persists only during the carrier lifetime (~2 ps) and returns to its original value as soon as the carriers recombine to disappear. Our data clearly demonstrates that magnetic properties, e.g., coercivity, can be strongly and reversibly modified in an ultrafast manner. We attribute the origin of this unusual phenomenon to carrier-mediated ferromagnetic exchange interactions between Mn ions. We discuss the dependence of data on the time delay, pump polarization, pump intensity, and sample temperature. Our observation opens up new possibilities for ultrafast optical manipulation of ferromagnetic order as well as providing a new avenue for studying the dynamics of long-range collective processes in strongly-correlated many-body systems. Finally, we propose an entirely nonthermal scheme for magnetization reversal which lead to new possibilities for extreme fast magneto-optical recording (>THz bit/sec).
Spatial correlations in time-resolved terahertz speckle patterns
We describe observations of the amplitude and phase of an electric field diffusing through a three dimensional random medium, using terahertz time-domain spectroscopy. These measurements are spatially resolved with a resolution smaller than the speckle spot size and temporally resolved with a resolution better than one optical cycle. By computing correlation functions between fields measured at different positions and with different temporal delays, it is possible to obtain information about individual scattering events experienced by the diffusing field. This represents a new method for characterizing a multiply scattered wave and could have potential applications to various disciplines such as medical imaging and coherent and incoherent communication.
Single-molecule Imaging and Manipulation
Nanostructures are of great importance for their scientific richness and for their potential to revolutionize critical technologies. Dramatic advances in the miniturization of mechanical and electromechanical structures have recently been made. We are striving to continue this innovation to the single molecule level. We have recently begun imaging and manipulating single molecules using the scanning tunneling microscope. A host of novel applications and new physics will be discovered as microelectromechanical structures are shrunk to the molecular-level.
Atomic-level Investigation of Fluorinated Carbon Nanotubes
Carbon nanotubes show great technological potential for their unique physical and electrical properties. There is currently a tremendous amount of research devoted to the fluorination and chemical modification of carbon nanotubes for both solvation and tuning of their electronic structure. Towards this end, the fluorination of single-walled carbon nanotubes was investigated by scanning tunneling microscopy (STM). The atomic-scale fluorine coverage on the carbon nanotubes was studied as a function of exposure time, fluorine gas concentration, and temperature using STM. The fluorinated areas on the nanotube appear as bright bands of various lengths with domain boundaries perpendicular to the tube axis. The electronic and chemical interaction between the fluorinated nanotubes and the substrate was also probed.
Finite difference time domain studies of optical properties of nanoshell structures
The optical properties of metallic nanoshell systems are investigated using the Finite Difference Time Domain (FDTD) method. The method provides a convenient approach for calculating several physical properties of nanoshells based structures including the optical absorption and scattering cross sections as well as the local electromagnetic fields near the nanoshell surfaces. The method is applied to silver and gold nanoshells and nanoshell dimers. Comparisons with classical Mie scattering are presented.
Surface Enhance Raman Scattering (SERS) and Surface Plasmon Resonance (SPR) Sensing with Metal Nanoshells
Metal nanoshells are nanoparticles consisting of a dielectric core surrounded by a metal shell, which exhibit strong, structurally tunable optical resonances. We investigated the application of metal nanoshells to surface plasmon resonance (SPR) sensing and surface enhanced Raman scattering (SERS). Traditional SPR sensors monitor changes in the surface plasmon excited on a gold film in response to a change in the refractive index at the surface of the film, due to the change in its local dielectric environment. We investigated the effects of SPR sensing by changing the refractive index of the medium surrounding nanoshells suspended in solution and by monitoring changes in the plasmon wavelength for a dilute coverage of nanoshells immobilized on a transparent substrate. Surface Enhanced Raman Scattering (SERS) is the phenomenon of Raman signal amplification observed when molecules are adsorbed on plasmon active surfaces. This effect can be used to improve the sensitivity of existing Raman based analytical techniques in biomedical applications. The surfaces of plasmon-resonant nanoshells possess an electromagnetic near field that can be precisely controlled through nanofabrication and easily quantified theoretically, making them ideal candidates for studying SERS. For the nonresonant molecule para-mercaptoaniline (pMA), the SERS response of a nanoshell film quantitatively matches the calculated electromagnetic response of a single nanoshell, when the nanoshell plasmon is tuned near the excitation and Raman shifted wavelengths. For SERS of resonant molecules, such as fluorophores, the pump laser is tuned to the HOMO-LUMO gap of the molecule, drastically increasing the Raman scattering cross-sections of the signals that overlap with the emission spectra of the molecule. The dependence of the enhancement on nanoshell plasmon resonance and fluorescent properties of the dye was studied, with special interest in the overlap of the fluorescence emission and Raman spectra.
Plasmonic Properties of Metallodielectric Periodic Structures
Recent investigations into the manipulation and focusing of optical fields by metal structures with subwavelength features have attracted enormous interest. Advanced nanofabrication techniques that realize subwavelength metal structures, the development of computational methods to analyze their electromagnetic properties, the observation of new phenomena such as surface enhanced Raman scattering, have all contributed to this resurgence of interest and the development of a new field newly dubbed "plasmonics". It is likely that the design and development of metallic nanostructures whose plasmonic properties can be tuned and manipulated will lead to the development of new optical components at the nanoscale, analogous to traditional optical components such as lenses, mirrors and waveguides. We report an experimental and theoretical study of the optical properties of metallodielectric gratings with subwavelength gaps in the thin metal limit. A mask-free method of fabrication for large area submicron silver gratings on silica substrates has been developed using soft-lithographic techniques. Two types of anomalies are found in the spectra of these gratings: an edge anomaly associated with the Rayleigh wavelength and a resonant anomaly associated with the excitations of surface plasmons. Measurements of the zeroth-order transmission and reflection have been performed to determine the spectral location of these anomalies and their dispersion relationships. The surface plasmons exhibit a well-defined gap in their dispersion relation, which is sensitive to the dielectric properties of the surrounding media.
Rice ECE Activities in the Medical Center: Multiphoton Microscopy and Spectroscopy
Multiphoton excitation utilizes infrared ultrafast (~100fs) laser pulses focused to very small (~1fL) volumes to create large instantaneous intensities sufficient to excite visible-excitation molecular fluorophores via the quasi-simultaneous absorption of two or more IR photons. This excitation mechanism provides two significant advantages over conventional "single photon" excitation: 1) IR light scatters much less than visible excitation, due to the ?-4 dependence of Rayleigh scattering, and 2) excitation is effectively confined to the focal plane, owing to the In dependence of multiphoton absorption, thereby reducing out-of-plane photobleaching and photodamage. These advantages make multiphoton microscopy particularly suited for imaging and spectroscopy applications in light-scattering tissue, such as living brain slices. Two Rice ECE projects in the Texas Medical Center center on the development of instrumentation employing multiphoton excitation. One project aims at the combination of multiphoton excitation with acousto-optic (AO) deflection, which allows a laser beam to be quickly (~1-10ms) re-positioned from any point in a microscope's field-of-view to any other. To achieve this, novel compensation approaches for both the spatial and temporal dispersion the ultrafast pulses incur upon propagation through the acousto-optic materials. Such an instrument would enable concurrent recording and stimulation of multiple sites in a living neuron, greatly enhancing the study of dendritic computation. A second project focuses on the characterization of a fluorescence correlation spectroscopy (FCS) apparatus aimed ultimately at the characterization of diffusion and binding of neural messenger molecules within living neurons. Work presented here demonstrates the robustness of the technique using control preparations as well as the optimization of the instrument's free optical parameters.
A Model of the Respiratory Central Pattern Generator in Rat
The respiratory central pattern generator (rCPG) is a collection of medullary respiratory neurons that generates the rhythm of respiration. The rCPG drives phrenic motor neurons (PMNs), which, in turn, activate the main muscle of respiration, the diaphragm. Despite extensive efforts, some deficiencies remain in existing models of the rCPG. We have developed a model of the rCPG based on experimental data from isolated medullary respiratory neurons. Our model mimics data from the respiratory medullary neurons and will be validated by comparing its normal activity and phase sensitivity to data obtained from the rat rCPG.
Vascular Smooth Muscle Relaxation: Modeling the Nitric Oxide/cGMP Pathway
In the this study, we present a model which extend our previous model by incorporating the NO/cGMP pathways. The model provides representation of the signal transduction flow of NO/cGMP pathway including NO sensing by sGC, sGC activated cGMP production from GTP, and cGMP-dependent intracellular calcium regulation as well as cGMP-modulated myosin light chain phosphorylation and muscle contraction. In comparison to experimental studies which usually take reductionistic approaches by investigating individual components of a system, our model presents an alternative to the study of NO related VSM relaxation mechanism as an integrated cell signaling system.

Last modified: October 7, 2003

Affiliates Meeting

Department of Electrical and Computer Engineering Rice University Affiliates Meeting - Poster Titles Wednesday, October 8, 2003

Department of Electrical and Computer Engineering
Rice University
Affiliates Meeting - Poster Titles
Wednesday, October 8, 2003